ANALYSIS OF NATIONAL PUBLIC RESEARCH FUNDING (PREF) … · 2020. 3. 12. · 6 Acknowledgements This...
Transcript of ANALYSIS OF NATIONAL PUBLIC RESEARCH FUNDING (PREF) … · 2020. 3. 12. · 6 Acknowledgements This...
-
ANALYSIS OF NATIONAL PUBLIC RESEARCH FUNDING (PREF) Final Report
Contract No. 154321
Emanuela Reale
2017
-
This publication is a Technical report by the Joint Research Centre (JRC), the European Commission’s science
and knowledge service. It aims to provide evidence-based scientific support to the European policymaking
process. The scientific output expressed does not imply a policy position of the European Commission. Neither
the European Commission nor any person acting on behalf of the Commission is responsible for the use that
might be made of this publication.
Contact information
Name: Koen Jonkers
Address: DG Joint Research Centre, European Commission
Email: [email protected]
Tel.: 00 32 2 295 7113
JRC Science Hub
https://ec.europa.eu/jrc
JRC107599
PDF ISBN 978-92-79-73407-6 doi:10.2760/19140
Luxembourg: Publications Office of the European Union, 2017
© European Union, 2017
The reuse of the document is authorised, provided the source is acknowledged and the original meaning or
message of the texts are not distorted. The European Commission shall not be held liable for any consequences
stemming from the reuse.
How to cite this report: Reale, E. (2017), Analysis of National Public Research Funding-PREF. Final Report,
doi:10.2760/19140
All images © European Union 2017
-
Contents
Foreword .............................................................................................................. 5
Acknowledgements ................................................................................................ 6
Abstract ............................................................................................................... 7
1 Introduction ...................................................................................................... 8
2 Government funding for R&D: previous experiences in analysis and data collection .. 10
2.1 Conceptualisation of R&D funding – the basis of data collection and analysis ..... 11
2.2 Building reliable data collection on project funding ......................................... 12
2.3 Other key sources of information on public R&D funding ................................. 14
3 Structure and methodological approach for the analysis ....................................... 16
3.1 Methodological approach - data collection ..................................................... 18
3.2 Data sources ............................................................................................. 21
3.3 Problematic issues in data collection ............................................................. 24
3.4 Data on EU structural funding ...................................................................... 26
4 Intensity of public funding for R&D in European countries ..................................... 27
4.1 Evolution of intensity of public R&D funding, 2000-2014 ................................. 27
4.2 Public/private sector performance of research, using public R&D funding .......... 30
5 Evolution of public R&D funding in European countries: institutional vs. project, policy objectives, beneficiaries ....................................................................................... 35
5.1 Characterisation of public research funding in the different countries ................ 35
5.1.1 The funding streams: project/institutional, recipients, research areas ........ 35
5.1.2 Characteristics of Project funding .......................................................... 41
6 Evolution in criteria and procedures for allocation of public research funding ........... 47
6.1 Proxy of competition in public R&D funding ................................................... 47
6.2 Performance-based funding ......................................................................... 49
6.3 Openness of funding to non-nationals ........................................................... 53
6.4 Delegation modes for project funding ........................................................... 53
7 Relevance of KETs and SGCs in national objectives and research priorities .............. 57
7.1 Key Enabling Technologies .......................................................................... 57
7.2 Societal Grand Challenges ........................................................................... 59
8 Transnational research funding.......................................................................... 61
8.1 Incoming European funding ......................................................................... 61
8.2 National funding to international performers and agencies .............................. 64
9 Funding bodies; evolution in management of funding mixes .................................. 65
10 Public research funding regimes in European countries ......................................... 76
10.1 Project vs. institutional funding .................................................................. 76
10.2 Performance based funding ....................................................................... 78
10.3 Managing agencies ................................................................................... 81
-
4
11 Conclusions and recommendations .................................................................... 84
12 References...................................................................................................... 87
13 List of abbreviations ......................................................................................... 91
List of figures ...................................................................................................... 92
List of tables ....................................................................................................... 94
Annexes ............................................................................................................. 95
-
5
Foreword
The JRC managed Public Research Funding (PREF) study has collected information on and
analyse national public research funding, providing an overview of its evolution and
current state in European and selected non-European countries, particularly as concerns
funding themes, fields and the types of allocation (competitive project based versus
institutional funding). More specifically, the study aims:
— For the countries considered, to collect quantitative and qualitative data (financial
data and descriptors) on the evolution of public R&D funding for the period 2000-
2014, including robust and reliable estimations of the shares allocated on project
basis and as institutional funding;
— To break down and provide a detailed analysis public R&D funding in terms of
competitive and non-competitive funding, and of the weight placed on scientific-
technological (S&T) fields, Key Enabling Technologies (KETs) and Societal Grand
Challenges (SGCs);
— To describe the funding allocation mechanisms, including flows to and from funding
agencies and the criteria used as the basis for allocation decisions.
— Using the above-described data, to analyse specific issues in public research policy,
particularly concerning the characteristics of national funding profiles and the funding
devoted to specific research objectives.
The study contributes to the Research and Innovation Observatory (RIO), and provides
support for development and implementation of evidence-based policies in allocation of
public research funding.
-
6
Acknowledgements
This Final Report summarizes the results coming from the PREF Project – Analysis of
Public Research Funding, funded by the European Commission (Contract No. 154321).
The PREF Consortium includes the following partners and teams:
— IRCRES-CNR – Research Institute on Sustainable Economic Growth, National
Research Council, Rome, IT (formerly CERIS-CNR) – Emanuela Reale (Coordinator),
Serena Fabrizio, Lucio Morettini, Emilia Primeri, Andrea Orazio Spinello, Antonio
Zinilli;
— USI – Center for Organisational Research, Università della Svizzera Italiana, Lugano,
CH – Benedetto Lepori;
— AIT – Austrian Institute of Technology, Vienna, AT – Matthias Weber, Barbara Heller-
Schuh, Thomas Scherngell, Georg Zahradnik;
— NIFU – Nordic Institute for Studies in Innovation, Research and Education, Oslo, NO –
Espen Solberg, Mats Benner, Ole Wiig, Lisa Scordato.
A number of national experts were also involved in the PREF project; they are:
— Ines Marinkovic, Elke Dall, Klaus Schuch from SNI - Centre for Social Innovation, AT;
— Jan van Steen from Rathenau, NL;
— Federica Rossi from Birkbeck University of London, UK;
— Daphne Getz from SNI- Samuel Neaman Institute, IL.
Elisabeth Pastor (OECD-NESTI WP) was the Quality Expert of the PREF project providing
expert advice for the methodology and data collection process.
The Final Report has 40 Public Funding Country Profiles (PFCPs) by different authors, and
the PREF Handbook for Data Collection and Indicator Production by Benedetto Lepori.
Serena Fabrizio elaborated the figures and tables based on EUROSTAT and OECD data,
and Andrea Orazio Spinello elaborated figures and tables based on PREF data, in the
Final Report and PFCPs. Serena Fabrizio and Andrea Orazio Spinello provided the editing
of the Final Report and the Annexes.
The PREF team acknowledges the collaboration from NSAs and Ministries of different
countries for the support given to the PREF data collection. The PFCPs indicate the
contacts in the different PREF countries.
Author
Emanuela Reale, IRCRES-CNR – Research Institute on Sustainable Economic Growth,
National Research Council, Rome, Italy
-
7
Abstract
This document presents the Final Report of the PREF study based on qualitative and
quantitative PREF data collection, and includes the suggestions and comments emerged
during the dissemination workshop held at Commission premises on March 2017.
The aim of the report is to analyse the national public funding for R&D in the 40 countries
under study, in terms of share of project versus institutional funding, mechanisms for
funding allocation, organisations for research funding, and extent of convergence around
the research themes and topics identified in FP7 and Horizon 2020. Some data on R&D
expenditures have been put in the Report as background information of the countries
analysed.
-
8
1 Introduction
The aim of the Public Research Funding (PREF) study is to collect information on and
analyse national public research funding, providing an overview of its evolution and
current state in European and selected non-European countries, particularly as concerns
funding themes, fields and the types of allocation (competitive project based versus
institutional funding). More specifically, the study aims:
— For the countries considered, to collect quantitative and qualitative data (financial
data and descriptors) on the evolution of public R&D funding for the period 2000-
2014, including robust and reliable estimations of the shares allocated on project
basis and as institutional funding;
— To break down and to analyse R&D funding in terms of competitive and non-
competitive funding, and of the weight placed on scientific-technological (S&T) fields,
Key Enabling Technologies (KETs) and Societal Grand Challenges (SGCs);
— To describe the funding allocation mechanisms, including flows to and from funding
agencies and the criteria used as the basis for allocation decisions.
— Using the above-described data, to analyse specific issues in public research policy,
particularly concerning the characteristics of national funding profiles and the funding
devoted to specific research objectives.
The study contributes to the Research and Innovation Observatory (RIO), and provides
support for development and implementation of evidence-based policies in allocation of
public research funding.
The current project is grounded in the tendering consortium’s previous research
experience in the study of public R&D project funding as part of the PRIME Network of
Excellence (EU Framework Programme 6), and in analytical work for JOREP (Investment
in Joint and Open Research Programmes) studies, funded by the European Commission.
From these experiences, the partners have gained greater capabilities in: a) developing
integrated conceptual frameworks for data collection on public R&D institutional funding
and project funding; b) examining policy intentions and results, to provide evidence
useful for design and implementation; c) integration of quantitative and qualitative
information on the characteristics of different funding streams, using sets of descriptors
that categorise the qualitative information for combined use with quantitative data.
This document presents the Final Report, which is the last deliverable of the PREF
Project. The report also includes the comments received during the dissemination
workshop organised at Commission premises and recommendations from the PREF core
team.
Using quantitative and qualitative data, the report analyses the evolution of public
funding for R&D in the countries under study in terms of: percentage of project versus
institutional funding; main allocation mechanisms for public funding; objectives of
research funded; convergence around the research themes and topics identified in FP7
and Horizon 2020. Individual Public Funding Country Profiles are included as annexes to
the report. These provide detailed information on national characteristics of public
funding for R&D and the evolution of allocation mechanisms.
The report presents a broad comparative analysis of the individual country profiles. The
countries under study are the EU-28 plus the accession countries and selected third
countries. The issues to be addressed in the analyses are:
— R&D funding intensity, based on GERD, HERD, GOVERD and GBARD statistics as
background information of the countries analysed;
— Characteristics of public R&D funding in terms of type of allocation mechanisms and
criteria (project funding and institutional funding);
— Proxies of competitiveness of public funding;
-
9
— Importance at national level of priorities identified in Horizon 2020, namely Key
Enabling Technologies (KETs) and Societal Grand Challenges (SGCs);
— Organisational characteristics of the funding bodies involved - Research Funding
Agencies (RFOs) and Umbrella Public Research Organisations (UPROs);
— Evolution of funding mixes and mechanisms used for allocation.
The Report is organised as follows. Section 2 briefly outlines the main literature
concerning previous relevant experience in data collection and analysis of national public
R&D funding. Section 3 presents the analytical framework and questions, summarises the
types of data available (EUROSTAT/OECD, other sources), and the methodological
approach for data collection under the current study. Section 4 examines the intensity of
R&D funding in the countries examined, in terms of overall dimensions, share of
government expenditure, share of GDP and total national research, as well as relative
channelling by NABS and FORD categories. This Section is mainly devoted to supply
background information on the countries analysed. Sections 5 and 6 analyses the
evolution of public funding types (institutional and project), the level of competitive
allocation in both types; Section 7 deals with the objectives and strategic research
priorities pursued under the country systems. Sections 8 and 9 examine the types of
funding bodies involved (RFO, UPRO) and the evolution of the funding mixes in the
different countries. Finally, Section 10 discusses the public funding regimes in European
countries, while the last section presents conclusions and identifies issues of data
availability and analysis still open to further examination.
The following Annexes are also included:
— Public Funding Country Profiles for each of the 40 nations studied, presenting data on
national funding devoted to R&D and summaries of the modes of coordinating public
funding. Country profiles follow the two templates used for data collection, in keeping
with PREF tender specifications. Full profiles are provided for EU-28 and accession
countries, while profiles for non-European countries (US, JP, CN, IL) and candidate
countries (FYROM, ME, RS, TR) are without analysis of funding instruments
(Deliverable D5.2);
— Handbook on PREF Data Collection, including explanations of indicator production, list
of S&T funding fields considered, funding categories by socio-economic objectives,
fields corresponding to FP7 thematic priorities, KETs and SGCs (Deliverable D4.1).
The structures of the main report and country profiles have been progressively refined on
the basis of the emerging data collection and availability, and the resulting real
possibilities for robust, confident analysis across the different countries.
-
10
2 Government funding for R&D: previous experiences in
analysis and data collection
Public spending for research and development is undoubtedly one of the most powerful
tools for government policy in areas of science, technology and innovation. On average,
public funding constitutes 35% of total R&D spending in European countries. However,
despite more than a decade of systematic study, advancement in policy knowledge,
coordination, and sharing of common spending targets, there is wide variation in
European countries both in terms of relative importance of public funding and the
mechanisms and criteria applied for its allocation.
Most European countries face strong fiscal constraints, and governments are increasingly
expected to provide greater accountability, transparency and effectiveness in public
spending. The expectations for research funding are certainly no exception. Hence, one
of the major challenges facing European policy on R&D is how to strengthen quality and
increase outputs given limited amounts of resources. Although several studies have
already sought to categorise and map the various national funding models (see for
instance OECD 2011), these models have changed significantly over the past few
decades, and many countries are currently reviewing and modifying their systems
(OECD, 2016). It is therefore worthwhile to provide an updated overview of funding
patterns and mechanisms currently in place in EU member countries.
Interest in potential variations of funding modes emerged in the 1970s and increased in
the subsequent decade, for two main reasons: the stagnation in volumes of public
research funding, corresponding with the contrasting emergence of new policy rationales
for efficient use of the funding (Lepori et al., 2007b). The conception was that
enhancement of competitive allocation mechanisms would stimulate better research
performance and more efficient use of resources, through selection of top research
groups, promotion of specific research themes and fields, supporting structural changes
in the means of knowledge production and application, competition and cooperation
between groups (Geuna, 2001; Braun, 2003). Many countries have embarked on reforms
in funding in response to new demands and opportunities. They have enhanced their
strategic-planning capacities and devoted more attention to the socio-economic
environment and evolving expectations of stakeholders. In 2003, an OECD study on
transformation of funding modes revealed that at that point, overall volumes of R&D
funding had increased, although public funding was generally increasing more slowly
than private (Maass, 2003; OECD, 2003).
In the 1990s, the trends toward competitive allocation gained strength. Government
funding increased for mission-oriented and contract-based research, which are conceived
as being more dependent on output and performance criteria. R&D statistics and
analyses indicated that nearly all countries intended to increase public funding, with the
increases concentrated in priority areas and new programs where funding would be on a
highly competitive basis. Also, competitive allocation of institutional funding emerged as
a central issue, but the capability to fully address a performance-based allocation faces
several constraints (Hicks, 2012). Despite this trend, there was a lack of data for
measuring these types of transformations, and for development of systematic
comparisons between countries and through time. The main reason for this was that R&D
statistics were designed to record expenditures at the overall levels of state allocations,
and then at the level of individual performers. In other words, the interest was in the
overall funding flows and research efforts, not the functional matters of how the
allocation was enacted (Lepori et al., 2007 b).
Moreover, an empirical analysis of the design of research programs in 34 European
countries (Optimat, 2005) showed that “national policies and programmes are being
developed without any obvious alignment with the parallel situation in other countries”.
Transnational research activities were still underdeveloped and there were many barriers
involving legal and organisational factors and research capacities. In other words, there
-
11
was little evidence of policy convergence, which refers to the tendency for countries to
grow more alike, particularly by developing integrated policy instruments for
transnational research (Knill, 2005).
Recently there has been increased emphasis on obtaining good statistics and indicators
of national public funding for R&D. The EC (2014) document, “Research and innovation
as sources of renewed growth”, describes the need to increase quality of public spending
“by allocating funding on a competitive basis, through open calls for proposals according
to excellence, for instance on the basis of international peer review, and by allocating
institutional funding on the basis of proven performance”. The document also reports on
European countries that are currently introducing greater competition in allocation of
research and innovation (R&I) funding. Monitoring and assessment of the evolution of
research in European countries is clearly very important. For this, it is necessary to have
quantitative data revealing the trends towards competitive and performance-based
funding, and qualitative information for characterisation of the structure and organisation
of the funding streams and instruments.
2.1 Conceptualisation of R&D funding – the basis of data collection
and analysis
During the years 2004-2008, the PRIME Network of Excellence, under the 6th Framework
Programme, conducted experimental research in analysis of public sector project funding.
This important step demonstrated the feasibility of developing new indicators of public
funding drawing on pre-existing data sources, and that these indicators could provide
useful results for comparative analysis of public research policies (Lepori et al., 2007a).
One of the major achievements was to operationalize definitions and methodologies for
collecting data on project funding. To this aim, a multi-layered conceptual framework of
government funding was developed, which allowed the shift from a performer-based to
agency-based approach in data collection. The new approach differs from the one
developed in the late nineties by RAND Europe (1999), which differentiated among
participants in the R&D system and among flows of funds. In the new approach, the
concept of “proximity to research performance” is used to classify participants (funders,
intermediaries and performers) and objectives (public or private), and the funding
arrangements are differentiated in terms of dimensions of control and competition.
The new conceptual approach began with recognition of the R&D system as a multi-
layered and multi-level space (Kuhlman, 2001; see also Rip, 1990) in which interaction
among four entities (policy, funding agencies, performing organisations, and research
groups) takes place. These entities represent different functions in research funding and
are organisationally separate, with some exceptions particularly the case of vertically
integrated national organisations that act as both funding agencies and research
performers, such as national academies of sciences in central and eastern Europe (Lepori
et al. 2009), and organisations like the CNRS in France (Thèves et al. 2007). The PRIME
project identified the wide use of two types of government funding mechanisms:
institutional and project funding. Institutional funding can be provided applying different
bases for calculation, and can be transferred as lump-sum or budget-line allocations, in
keeping with the level of autonomy granted to the research performers (Jongbloed,
2001). In the past, such funding generally came without conditions. More recently there
has been a clear trend towards inserting performance or quality-based mechanisms, to
enhance the accountability of organisations and stimulate competitive behaviours.
Institutional funding can also be routed differently to different types of performers. For
example, some countries apply “dual system” funding for HEIs, meaning for both the
institutions and their individual researchers and research groups. On the other hand,
governments generally do not provide dual funding for public research organisations.
Allocations of project funding are generally based on competitive processes, on the basis
of “bid” applications submitted in response to calls for tenders and notices issued by a
funding agency, with evaluation using different types of peer-review process. Project
funding can also be contract based, with specific objectives and milestones. Project
-
12
funding has increased substantially in a number of countries in the past decade, including
for projects with joint government and industry support, as well as involving
governments and agencies from different countries.
Scholars focusing on science policy have also developed a very specific conception of
“delegation models” in public funding (Braun, 2003). This approach interprets the
relationship between state and scientists as one of delegation and then tries to
characterise funding policies and instruments in terms of their underlying delegation
models. The current PREF analysis links the instruments used to allocate public research
funding to the abstract categories of delegation modes, attempting to reveal the general
features of relationships between the state and national science systems. An exploration
of this approach-using project funding data (Potì and Reale, 2007), demonstrated that
availability of both qualitative and quantitative information is crucial to such policy
analysis.
In terms of general strategies for developing indicators and applying them in policy
analyses, the PRIME experimental activity yielded a number of relevant lessons (Lepori et
al., 2007b). First, the development of indicators (and data collection) should be driven by
the underlying policy questions, related to program objectives and intended impacts.
Second, the resort to reasonable simplifications is key to the successful development of
indicators. This means collecting data on the programs and characteristics that are truly
relevant to the underlying questions, as well as using proxies and estimations where
there is reasonable confidence that these can provide a correct assessment. Third, when
developing indicators in new fields, rather than giving all the definitions and categories
“ex ante”, an interactive approach will often be more successful. Starting from fairly
general categories, and progressively refining them based on experience in data
collection and analysis, assists in getting the right indicators and making data collection
feasible. Fourth, careful consideration must be given to the choice of data sources, the
means of combining them, and the approaches to actually collecting the data, taking into
account availability, problems in quality, as well as the efforts necessary for collection
and “cleaning”.
2.2 Building reliable data collection on project funding
The PRIME approach was further developed under a NESTI (OECD National Experts on
Science and Technology) pilot project examining government budget allocations for R&D
(GBARD). The project aimed at integrating a small core of indicators on project funding
into the official statistics produced by the various national services. GBARD data
collection is not confined to national statistical offices, and the project extended to other
national institutes recording information in this area. The aim of the experiment was “to
propose a methodology for internationally comparable indicators on modes of public
funding; collecting such indicators and assessing the feasibility of extending their scope
by developing guidelines for the implementation of the methodology as part of the
regular indicator activities of OECD and other organisations like EUROSTAT” (van Steen,
2012).
The NESTI project involved two rounds of data collection in a total of 18 countries. It
demonstrated that, at least for these countries, national data on appropriations and
outlays can serve as useful sources for identifying and distinguishing allocations to
institutional and project funding, although there can be problems of availability (Steen,
2012). Complementary qualitative data is required to interpret the results, since the
quantitative data are insufficient for comparative reasoning, given differences in country
R&D funding policies, funding systems, mechanisms and practices. Knowledge in all
these aspects is essential for policy analyses; therefore, data collection must be designed
to detect such differences at country level. One of the most promising results of the
OECD experiment was the reliability of the outcome in terms of data produced, a result
demonstrating that the achievement of basic levels of international comparability is an
attainable objective. At the moment, data across the OECD nations are not fully
comparable, however, EUROSTAT has adopted the project/institutional funding as an
-
13
“optional” variable for reporting. The NESTI experiment also provided important input for
revision of the Frascati Manual on practices for surveys concerning R&D, for example
introducing the notion “funding agencies”, and their relative investigation.
Data collection on public funding for transnational research
The JOREP project (Investments in Joint and Open R&D Programmes) provided further
progress in identifying indicators and data for analysis of public funding. The project
covered 11 European countries, collecting descriptors aimed at characterizing the
features of the different transnational programs, the related funding agencies, and data
on funding, modes of allocation and budget management.
The development of transnational public research began as early as the 1950s. Both the
form and extent of cooperation have evolved significantly over recent decades, giving
rise to a highly differentiated landscape of programs and agencies. Broadly speaking, we
can identify a first phase in which the main cooperation mechanism involved creation of
intergovernmental research facilities in fields where the scale of investment requires that
the European countries joined forces. Examples are CERN in nuclear physics, the
European Southern Observatory in astronomy, and the Joint Research Centre of the
European Communities in the nuclear sector. In the sectors of nuclear energy (Euratom)
and space (European Space Agency), the main function assumed by intergovernmental
bodies was as funding agencies, promoting research in member states through financing
of projects and contracts. Euratom has since been integrated into the EU framework
programmes, but ESA remains by far the largest intergovernmental funding agency in
Europe.
In the second phase, beginning in the 1970s, the focus has largely shifted to creating
intergovernmental funding schemes for cooperative research, to be realised through
networks of national performers. This has involved moving away from creating large-
scale facilities towards promoting cooperation by European actors, and establishing
critical mass through joint projects. Two broad types of intergovernmental programs
have emerged: i) programs coordinated by a European agency, but with funding
managed through national agencies (mostly à la carte), and thus without trans-border
flows COST (European Cooperation in Science and Technology, the EUREKA network
(launched 1985) and most European Science Foundation schemes belong to this
category; ii) programs where both the management and the funding are handled through
a European agency e.g. European Space Agency, and the very obvious cases of the
European Framework programmes.
JOREP recommended an approach to data collection based on three critical components:
definition of the dataset perimeter, the descriptors, and obtainment of data on funding
flows. The construction of a list of joint programs is central to data collection. The list
should provide the following information: program name, financial data, participating
countries, and participating agencies within each country. The set of descriptors
developed under the project proved highly useful in analysing the landscape of joint
programs over a period of several years. Given the fact that the organisational
characteristics of joint programs remain reasonably stable, the descriptors allowed
observation of the landscape changes, the transformations in program characteristics,
and the related changes in participating countries.
At the time of the JOREP project, GBARD data had not yet been demonstrated as useful
in monitoring funding under joint programs. Sufficient information was at times
unavailable from national statistical offices. Instead, expertise was developed in
retrieving data on funding flows using web-based sources and by direct collection of
information from national funding agencies. Also, rather than the breakdown of
beneficiaries described by the Frascati sectors, JOREP recommended provision of a
simple breakdown of public and private beneficiaries, as more doable and relevant to
understand how the program truly functions (Reale, Lepori, et al., 2013; Lepori, Reale,
Laredo, 2014; IUS, 2014).
-
14
The JOREP project made important steps in data collection, since it showed how to merge
budget data with data useful in identifying the actors’ roles and the funders’ institutional
strategies. Moreover, the project developed techniques for investigating the levels of
national integration in joint R&D programs, relative to the characteristics of the individual
programs. This provides an empirical basis for representation of the resources mobilised
and the opportunities achieved in the areas the policy makers originally intended for the
program and funding schemes (Lepori, Reale, Laredo, 2014; Reale, Nedeva, Thomas,
Primeri, 2014).
2.3 Other key sources of information on public R&D funding
ERAWATCH was the European Commission's information platform on European, national
and regional research and innovation systems and policies, created in 2005. Its main
objectives were to support policy-making in the research and innovation field in Europe
and to contribute to the realisation of the European Research Area (ERA). ERAWATCH
has been the core of the EC information service on R&I, and a unique source of policy
intelligence. The launch of the platform benefited from the previous experience of the
European Trend Charts on Innovation, and continues to focus on production of yearly
country reports and other national reporting templates, some of which contribute to
describing the funding agencies and support measures within the different countries.
However, ERAWATCH was not limited to the national level, and also gathered and
analysed information on regional and European level R&D policies, actors, organisations
and programs. The information and analyses were designed to serve policy makers and
managers, particularly to assist in identifying policy options and improve the coordination
of scientific and technological activities.
ERAWATCH has been discontinued but its experience has been further developed in the
Joint Research Centre's (JRC) Research and Innovation Observatory (RIO)
(https://rio.jrc.ec.europa.eu), which is a new EC instrument to support member states
and countries associated with Horizon 2020 in policy design and implementation, as well
as for assessment of reforms. RIO makes major efforts to understand changes in the
allocation of public and private funding, with particular interest in the emergence of
competitive allocation and performance-based allocation modes on the public side
(Jonkers & Zacharewicz, 2016).
Project funding is also used for promotion of research in priority areas, and to pursue
other specific public policy objectives (e.g. search for excellence, internationalisation)
and strategic interests (key enabling technologies, research areas linked to “grand
challenges”). An OECD report presents the interesting experience of surveys intended to
obtain data and evidence on how governments fund and steer public research in higher
education and research institutions through research excellence initiatives (REIs), as a
new form of competitive research funding. Three surveys were carried out, reporting on
56 schemes in 20 countries (OECD, 2014). The first survey addressed the government
agencies responsible for administering REI funding to higher education and research
institutions, seeking to define the characteristics differentiating REIs from other funding
streams. The second survey addressed centres of excellence funded by REIs,
investigating their management structure, funding schemes, measurements of impact
and sustainability, cooperation with public and private sectors, and perceived long-term
effects of their research. The final survey addressed the institutions hosting the centres
of excellence, inquiring into their administrative arrangements, financial and research
objectives, and the impact of the REI-funded centres on the larger institutions.
The above experiences in analysis of public funding provide constructive evidence and
suggestions for collection of qualitative and quantitative data on “framework” competitive
funding schemes, where several instruments contribute to a specific policy objective,
mobilising important volumes of resources in several countries towards what is
recognised as a strategic goal, capable of impacting the effectiveness of R&D investment
and competitiveness of the economies.
-
15
This concludes our brief section on the background of government funding for R&D, and
relevant experiences in data collection. The possibility of assessing the overall trends of
research policies in the European Union countries, and benchmarking policy
developments in the EU against those taking place in key trading partners, clearly
emerge as important issues. Such capacities create strong evidence base, useful in
creating decisional advantages for the Commission, the EU member states and
stakeholders.
This report on the “EC PREF project” now proceeds to analyse public research funding,
building on past experience and using original collection of quantitative (financial
amounts) and qualitative (organisational descriptors) data.
-
16
3 Structure and methodological approach for the analysis
Funding is one of the major instruments accessible to governments for steering science
and innovation systems (OECD, 2011). Many countries have introduced reforms of their
R&D funding systems, to foster excellence, knowledge transfer and socio-economic
innovation, in which the changes typically reduce the proportion of institutional funding in
favour of more selective and competitive systems. This trend has been investigated,
however the indicators selected for the purpose (see review in section 1) have only been
able to partially depict the importance and extension of the two key types of allocation
systems: institutional funding, where the majority of resources are allocated directly to
institutions according to particular arrangements (e.g. historically based, application of
formulae, reference to performance indicators, budget negotiations between actors), and
project-based systems, where scientists compete to obtain funding from external
sources.
The debate about the various consequences of the shift toward project-based
arrangements is still open. Different evidence has been provided in the literature about
whether the increasing reliance on competitive funding at the expense of block grant and
long-term institutional funding has resulted in an emphasis on short-term low-risk
projects at the expense of long-term fundamental research. Other concerns include the
extent to which program-oriented funding might hinder possibilities of exploring non-
priority areas, or how conditions that generate markets and demand from research
users, which are not the same for all scientific and technological areas, could push
certain fields while depriving others of vital resources (OECD, 2011).
Despite such debates, the recent policy statements concerning the establishment of the
European Research Area stress the importance of making national-level funding systems
more effective, and indicate the importance of competition towards this aim. National
strategies towards building an integrated European research and technology area have to
be improved, and national research agendas and strategies must be better aligned with
one another within this overall area. One common feature of the Member States’
strategies is support for further implementation of project-based funding, which is now
observed in 21 countries (EC, 2015). Moreover, the mechanisms for project funding
allocation show a shared trend towards complying with high standards of peer review.
Indeed, this emerged as a condition for agreements between the funding agencies
operating in different countries, towards engagement in transnational cooperative
research programs (Reale et al., 2013).
Nonetheless, a major difficulty in investigating the different national R&D funding
systems is the insufficient availability of quantitative and qualitative data on the
instruments and actors involved in their management: the specific research funding
organisations and/or “umbrella public research organisations”, which in the latter case
act as funding bodies for their sub-organisations and other providers. The current project
for development and analysis of data is intended to contribute to better knowledge on
the characteristics of national public funding for research, including the priorities,
instruments, actors and their strategies.
The questions to be addressed by the PREF analysis devolve from the main problems, as
described above:
— How is public R&D funding evolving in the countries considered (based on data
collection over the period 2000-2014)?
— What sectors of performers are funded in the different countries?
— What is the evolution of institutional and project funding allocation mechanisms in the
different countries?
— To what extent is the trend towards competitive research funding observed in the
policies of European and non-European countries?
-
17
— What are the characteristics of the funding mixes in the considered countries?
— What policy objectives characterise the R&D funding in the countries under study, and
how do the different actors allocate the funding?
— Does public research funding show evidence of converging patterns among European
nations, as regards priorities and type of instruments?
The report elaborates comparisons of public research funding between countries.
Qualitative information, integrated and checked by empirical data, is used to develop an
understanding of the characteristics of national systems, including the relationships
between government, funding agencies and research actors. The report:
— Identifies evolving patterns of institutional and project funding modes, comparing
ERA and non-ERA countries, and countries in western and eastern areas of Europe;
— Maps public funding characteristics by country, to understand what they reveal of
national policy and objectives, particularly in the area of public-private collaboration;
— Identifies and compares institutional arrangements, by country (project funding
versus core funding configurations, vertically integrated systems);
— Assesses the extent to which national funding instruments integrate European-level
strategic objectives, particularly the “cross cutting” Key Enabling Technologies (EC,
2012) and Societal Grand Challenges, defined under Horizon 2020.1
The analysis takes a comparative approach, examining EU-28 countries, candidate
countries (FYROM, Montenegro, Serbia and Turkey) and non-European countries (China,
Japan, Israel, US) as far as data are available. The comparisons are based on the
national data and information presented in the country profiles (Annexes 1 to 40). Within
the EU-28 countries, special attention is devoted to addressing the characteristics of
national R&D funding in the eastern, western, northern and southern areas of the
European whole.
From a theoretical point of view, the report builds the analysis around examination of
coordination modes for research funding, which is the approach best adapted to
examination of country differences. Under this perspective, national funding systems are
composed of combinations of organisational forms, presenting different characteristics
and conditions. The different funding mechanisms generally correspond to coordination
modes, which determine the requirements for actors and their means of interaction for
achievement of collective action, and scientific, economic and social impacts. In other
words, different agents move within the specific interaction spaces of national systems,
in correspondence with the different coordination modes for public funding: project-
based, mixed, or vertically integrated (Lepori, 2011).
In the project-based mode, we can expect interaction spaces where different research
funding organisations (RFOs) coexist. State controls could range from partitioning of
funds between the different RFOs to determining allocation criteria, but in any case, the
RFO actors retain control over selection of beneficiaries. In contrast, the vertically
integrated mode is dominated by: i) “umbrella organisations”, which serve as layers for
coordination of funding on the basis of extensive state delegation. Mixed systems feature
combinations of project funding and other modes, such as allocations to consortia,
networks, or “centres of excellence”, which then regulate the further division of resources
among partners; ii) higher education core funding, where the main coordination
relationships are between the state and HEIs. The balancing between different modes
configures the national research systems and thus influences the European nations’ joint
programming of national and transnational research.
1 http://ec.europa.eu/programmes/horizon2020/en/h2020-sections
-
18
3.1 Methodological approach - data collection
The current project takes a unified approach to collection of quantitative and qualitative
data on national public research funding, building on a conceptualisation of national
systems successfully applied in previous data collection on project funding (Lepori et al.,
2007). National systems are conceived as consisting of four layers. The first is the state
layer, meaning nation and European Union levels. These are the levels at which the
overall volume of budgetary appropriations for R&D activities is decided, as well as their
division in broad streams. The second layer is composed of “research funding
organisations” (RFOs), a broader term than “research funding agencies”, meaning all
organisational entities assigned by the state to distribute money to research performers.
When responsibilities for managing funding are conducted by ministries, this term refers
only to the units specifically in charge of distributing funding. The third layer is composed
of large, stable “research organisations”, encompassing diverse research topics, with
operations by various research groups. The fourth layer is composed of the individual
research units and groups, of different types, representing the true scientific “production
units” (Etzkowitz 2003; Joly and Mangematin 1996).
In some countries, we can also find very large “umbrella public research organisations”
(UPROs), with dual functions of both managing extensive parts of the national research
system and acting as RFOs for their own laboratories and research groups. Examples
would be the CNRS of France, CNR of Italy, and CSIC of Spain (Thèves, Lepori and
Larédo 2007).
The analysis considers two main allocation mechanisms for public funding: institutional
and project-based. This distinction has long been known in research studies, but has only
recently been operationalized in quantitative analyses (Lepori, Dinges, Poti, Reale,
Slipersaeter, Theves and Van den Besselaar 2007; Steen, 2012). The conceptual
definitions of project and institutional funding are provided in the PREF Handbook (p. 9-
10):
Project funding is defined as money attributed to a group or an individual to perform a
research activity that is limited in scope, budget and time. It is identified based on three
main characteristics: a) funds are attributed directly to research groups and not to whole
organisations; b) the scope and duration of research supported are limited; c) funds are
attributed by an RFO external to the organisations of the end research groups. “National”
project funding is defined as national public appropriations allocated through project
funding.
Institutional funding is funding attributed to research organisations (PROs, HEIs) for
ongoing activities, usually over unrestricted periods of time. The amount of funding may
vary on a yearly basis, but is generally not earmarked for specific activities or
organisational subunits. Internal allocation is left to the performing organisation. In most
countries of Europe, block transfers to universities are allocated as lump sum institutional
funding, and comprise the large part of the individual institutional budgets.
Also, the Handbook provides the conceptual definitions of Research funding Agency
(RFO) and Umbrella Public Research Organization (UPRO) as main actors on the funding
side (p. 10-11):
Research funding organizations are organizational entities that distribute public funding
for research on behalf of the State. The definition adopted is extensive concerning the
legal status and the position in respect of the State, covering both independent agencies
at arm’s length from the public administration, like research councils and ministries, and
offices within the public administration, which perform this role. Most research funding
organizations distribute project funding, but in some countries, RFOs (like higher
education councils) are also in charge of distributing institutional funding. In a few cases,
both functions are present, like in the case of research councils managing national
facilities.
-
19
Umbrella Public Research Organizations are national-level organizations with the mission
of organizing research activities in a specific domain. Unlike normal PROs, they mostly
host research laboratories distributed over the whole national territory and they are
delegated by the State to manage a specific field of national research policy. Umbrella
PROs in many cases have a dual function, i.e. to directly manage laboratories and
scientist’s careers on the one hand, and to provide competitive projects funds on the
other hand.
The study methodology draws on the experience of the PREF consortium participants in
previous projects, and is based on the following principles.
1. The design of data collection, both quantitative and qualitative, derives from the
conceptual scheme illustrated in Fig. 3.1. The scheme identifies the key actors and
main funding streams of the system. For each component, we identify the available
data sources and organisation of data collection. Information about the public
availability of the data are also supplied. Once data for the different parts of the
system have been produced they are integrated to yield aggregate indicators
characterizing the national systems, which can then also be used in systematic cross-
country comparisons. Broad national aggregates and disaggregated data are
systematically reconciled to improve the robustness of the data.
2. The main focus of the report is on providing detailed analysis of “policy intentions”
and the different mechanisms through which research funds are allocated to
performers. The inquiry into the actual research performed is limited to broad
research types, topics and the relative volumes of funding, and is only carried to the
extent that it serves in analysing public policies. This focus conforms to the goals of
the PREF project tender, but is also suited to data availability, since the funding
streams are more aggregated at the policy-making and RFO/UPRO levels and hence
easier to quantify than R&D expenditures at the detailed, granular level of
performers.
3. Standardised descriptors are developed, based on closed sets of categories,
concerning funding policies, organisational structure, selection criteria and allocation
modes. This approach establishes a data collection task that is reasonable in scope,
and improves comparability of the information collected. It also allows statistical
testing of associations between program characteristics, which would be problematic
using purely qualitative descriptions (Lepori, Reale and Larédo 2014). The approach is
particularly suitable given the large number of units being compared, particularly at
the program and RFO levels.
4. The study examines only the “larger” funding organisations, programs and streams.
In practical terms this means limiting data collection to funding streams representing
at least 5% of total government budget allocations for R&D, according to EUROSTAT
statistics. The residual amounts, if they represent consistent amounts of public
research funding, are viewed as aggregate funding. This solution is supported by
empirical evidence from the JOREP project, of strong concentrations of funding in
small numbers of streams and agencies, while the rest is divided between large
numbers of small RFOs (Reale et al., 2013). The Handbook provides for flexibility in
singling out RFO streams below the threshold when they are relevant at national
level. The expert-based identification of the funding streams to be included in the
study is thus a core aspect of methodology.
The components of data collection methodology deal with the different parts of a national
research funding system: public appropriations of funding for research; public R&D
procurement; incoming EU funding; characterisation and data on research funding
organisations; characterisation and data on research funding instruments;
characterisation of umbrella research organisations; analysis of R&D execution.
In this respect, it is useful to recall the general characteristics (and limitations) of GBARD
data, which apply by definition also to the PREF data collection. A reference can be made
to Chapter 12 of the OECD Frascati Manual (OECD, 2015). GBARD is based on budgetary
-
20
information (budget provisions) and not to actual expenditures. The calculation is based
on the identification of all the budget items related to R&D, and on measurement or
estimation of the funding related to R&D contents. Government perimeter includes
funding coming from central/federal government, and regional/state government when
funding is relevant. Structural funding therefore is not included as well as indirect
funding and funding from public enterprises. Since GBARD is constructed differently in
different countries, this may lead to limitations in the comparability of the findings.
According to the Frascati Manual (OECD, 2015) GBARD covers government-financed R&D
performed in government establishments and government-financed R&D in the other
three national sectors (business enterprise, private non-profit, higher education) as well
as abroad (including international organisations). The main advantage of GBARD is that it
allows measuring the objectives of public R&D funding, which is relevant information to
analyse the evolution and the characteristics of the government R&D policies.
The full methodology, including conceptual scheme, definitions, classifications, indicators,
variables and data sources, is presented in the PREF Handbook on data collection and
indicator production. The reference perimeter is government budget allocations for R&D.
Data on national GBARD, including their breakdown by NABS categories, are derived
from EUROSTAT and OECD databases, and are provided in the report as a reference.
National public funding is distinguished at two levels:
— Major “funding streams” within the national GBARD, such as core allocations to higher
education institutions, allocations to funding agencies and large PROs. The PREF
report provides the basic characterisation of these funding streams.
— Concerning project funding, the more fine-grained level of “funding instruments”,
using descriptors to identify the main characteristics of government operation. The
funding instruments are in turn connected to performing sectors. A given type of
funding instrument is not necessarily unique to an individual national program; the
same instruments can be used in different programs. The level of granularity in
reporting depends on the national funding structure, and in some cases on the
capacity to disaggregate the different streams. As a rule, a funding instrument is an
articulation of a stream which groups programs of similar structures and objectives.
In this report, figures related to i) project funding by beneficiary and type of funding,
ii) public private-cooperation, iii) institutional funding, and iv) GBARD for Key
Enabling Technologies (KETs) and Societal Grand Challenges (SGCs), are calculated
at the instrument level. This means that the amounts can be different from the
figures calculated at the stream level, because streams are more complete, and also
include state funding to international agencies.
Funding streams and funding instruments are linked to the research funding
organisations that manage them, hence revealing the bridges between funding streams
and their respective managing organisations. Fig. 3.1 shows the basic data structure.
Detailed information about the process of data collection have been provided in
deliverable D2.4 and Annexes of the PREF project.
-
21
Figure 3.1 PREF data structure
Source: PREF Handbook, p. 33
3.2 Data sources
The analytical report uses data from the following sources:
— EUROSTAT/OECD data on gross domestic expenditures for R&D (GERD), GERD funded
by government (GOVERD), GBARD, business expenditures for R&D (BERD) funded by
the government sector, and higher education research and development (HERD), as
collected in the First Data Package (Deliverables 2.1). The Report uses the data
retrieved from EUROSTAT on April 2016.
— Data from quantitative and qualitative PREF data collection at the national level
(Second Data Package, Deliverable 2.4). Data on funding from international research
agencies to national performers are also included.
Although the PREF analyses use official statistics on funding and expenditures, they do
not attempt to reconcile the total volume of funding/expenditures across levels, therefore
the totals by level might be slightly different. Comparability between countries is still
assured by calculation of the following data, for each country and year:
— Total GBARD, from R&D statistics.
— Total volume of all funding streams included in data collection (with identification of
the funding sources in the relevant countries).
— Total volume of the funding instruments included in data collection.
— Total R&D expenditures funded by the government sector.
Totals 1 and 2 should generally be very similar.2 The total at the instruments level should
be near the total GBARD, less the amount of funding streams transferring funds to RFOs
and performers abroad, and should be quite similar to total R&D funded by government.3
2 Differences might come from the inclusion of exchange funds, which are currently not always covered by
GBARD. 3 Difference might arise by the fact that funding instrument might also include some PNP or Business Enterprise
funding.
-
22
The different data from official statistics and PREF data collection serve as
complementary sources for mapping the intensity of public R&D funding, and for
deepening knowledge of the types of allocation and competitiveness, as well as the
responsiveness of countries to emerging R&D priorities at European level.
The issue of public support through tax incentives
Over the past two decades, an increasing number of countries have introduced schemes
to promote business R&D through tax incentives. Technically, such indirect support is not
considered as public funding, but it in fact constitutes an important part of the policy mix
and portfolio of public support for R&D. In many countries, there has been a trade-off
between public funding and such indirect support, sometimes involving a remarkable
decline in direct funding following establishment of the tax incentives.
Some countries have started to include the calculation of foregone tax revenues in their
total amount of public R&D funding. According to the OECD (2013), 27 out of 35 member
countries are now practicing some form of R&D tax incentives as part of public support to
R&D. However, countries vary widely in terms of the relative importance of these
incentives as a share of total public support to business R&D, from above 87 per cent in
the Netherlands to only 8 per cent in Italy.
As a result of the widespread use of R&D tax-incentives, the OECD has started to collect
harmonised data on the scope, profile and size of national tax-incentives. However, these
measurements are still considered experimental in nature. The current report does not
describe or analyse tax incentives as part of overall examination of national public
funding mechanisms and agencies, since the mandate did not propose inclusion of this
dimension, and due to the limitations of existing data collection and analysis. However,
despite the experimental nature of the OECD work, it is interesting to look at the
available data. These appear to show that in some European countries indirect funding
has a very important role in the national R&D strategy. One example is France, where
the volume of tax incentives is higher than direct government funding of business R&D,
measured as percentages of GDP (Figure 2.1).
Figure 3.2 Direct government funding and tax incentives for business R&D: percentages of GDP (2013)
Design and sources for qualitative analysis
The use of qualitative data in the descriptors permits analysis of the evolution of public
funding in terms of the characteristics of the instruments, which in turn reveal the policy
priorities and the intentions of the different actors.
The categories of qualitative information collected are those that serve in characterising
the funding streams and instruments. The data include that from desk research on recent
developments in funding mechanisms among the countries of the PREF partners. This
research complements the development of the descriptors foreseen in the data collection,
by integrating other information useful for the data analysis and for the country profiles.
In this respect, the most important source is the ERAWATCH database, followed by
OECD, European Science Foundation (ESF) and individual country sources.
-
23
The combination of different sources aims at gathering contextual and systemic
information for each country as background for understanding variations in R&D funding
structures. In addition, the network of national experts plays a role in filling in or
explaining gaps in data. The table below (Table 3.1) summarises the main categories of
indicators.
Table 3.1 Main categories of indicators
Source: PREF.
-
24
3.3 Problematic issues in data collection
The consistency of the data in the current report with that from official statistics is very
good. In particular the total GBARD by country at stream level is consistent with total
GBARD from EUROSTAT statistics. Moreover, the quality checks during data collection
guarantee very good comparability of data between countries (see Deliverable D2.4 for
details on quality checking and data quality).
In general, there is also good consistency of the PREF indicator on project funding
allocation with EUROSTAT data, for the limited number of countries that report to
EUROSTAT on this funding division. However, the consistency is not good for
Luxembourg, Poland and Portugal, where the EUROSTAT calculations show different
percentages of project funding out of GBARD compared to PREF data. The PREF method
of checking with experts and NSAs in different countries indicated greater reliability in
PREF data collection compared to that from EUROSTAT.
However, there were also problematic issues in data collection, one of which was the
strongly varying responsiveness of national contact points, particularly the NSAs. Both
deliverable D2.4 (Table 1), and the PREF database metadata provide more detailed
reports on all the country-specific problems. Under points 1 to 4, we limit ourselves to
grouping the countries relative to the different types of problems, and describing how
these were managed for the study.
1. Countries providing no or limited data
Despite determined efforts, there were four countries for which the only data obtained on
public R&D funding were those from official statistics. These were:
China. The first approaches for data collection received little response and led only to
general information and links to Chinese S&I policy documents. However, from this
beginning, we were able to retrieve data for elaboration of an overview of the major
RFOs and funding streams. The more difficult task was to retrieve official amounts
describing the allocations from ministries through single RFOs and funding streams. This
challenge becomes still more difficult since China recently initiated an extensive reform of
its S&T system. Under the reforms, about one hundred national S&T programs managed
by a mix of ministries and departments will be classified into five categories. More
importantly, the ministries and departments will then no longer manage specific projects
directly. This has implications for the PREF data on structure of RFOs and funding
streams, as collected for the current study.
Japan. The first approaches for data collection received little response. One problem
seems to be that a range of individual ministries holds strong positions in the funding
system, and that there are few persons able to provide consistent funding data across
ministries. There is also little material in English, from which to extract an overall picture
of funding. It was particularly difficult to link system descriptions with the corresponding
funding data over several years. For all streams, data distinguishing between project and
institutional remain uncertain.
Turkey. For this country, it was not possible to receive data on amounts for funding
streams. Cooperation from national authorities responsible for R&D statistics was very
limited. Publicly available sources proved useful for obtaining aggregate data on
government R&D expenditures, but little else.
Latvia. Despite a number of approaches by e-mail and telephone, the NSA did not send
any information or data about public R&D funding. Data on RFOs and funding streams
were developed based on ERAWATCH/RIO reports and external evaluations of the Latvian
R&D system, but obtaining data on the evolution of funding streams and instruments
over several years remained difficult. We compiled the data on funding agencies, streams
and instruments available from external evaluations and obtained a quality check and
supplementary data from non-NSA contacts in Latvia, but the analytical results should
still be considered with caution.
-
25
2. Countries lacking robust data
For some countries, primarily in Eastern Europe, there are problems of robustness and
consistency: Bulgaria, Croatia, Estonia, FYROM, Montenegro, Hungary, Serbia, and also
to some extent Luxembourg and Romania, where several types of data were missing.
The different problems and solutions adopted are detailed in the country profiles. In the
current report, we analyse and discuss only those data that are robust enough to
guarantee good comparability with other countries, which mainly concern the stream
level. Specific details and considerations on data availability are indicated in the text and
the notes to charts.
3. Countries with limited time series
The goal of the current report was to obtain data for the period 2000 to 2014. However,
for countries where GBARD data collection is conducted at two-year intervals, or where
NSAs had not yet released data for 2014, the last year reported may be 2013. In some
cases, there were also problems in availability of the full-time series:
— Lack of continuity because the methodology of GBARD data collection changed, most
frequently affecting data from 2000 to 2006, sometimes up to 2008;
— Because of lack of resources, some NSAs have not collected data on instruments, and
data disaggregated at the level of funding streams. Elaborating this sometimes
requires difficult work to build up categories of streams, particularly when reliable
alternative sources (e.g. annual reports of RFOs) are unavailable. Where possible,
estimations were used to reduce these shortcomings (see public funding country
profiles for details). These were always checked with the relevant NSA prior to use in
the analyses and report.
— Problems with GBARD data collection and availability at the national level, impeding
development of full time series.
These problems are presented in some detail in the country profiles and the PREF
database metadata (methodological notes).
4. Problems in breaking down data by fields, classifications and beneficiaries
A final set of limitations concern difficulties in breaking down the data by detailed fields.
For both institutional and project funding, it is impossible to consistently break down the
data on public allocations (GBARD) by field of R&D (FORD), either because the allocation
is in the form of core funding to specific institutions, who then serve as final actors in
deciding distribution, or because the possibility to disaggregate a project funding
instrument into separate fields is possible only at the level of execution.
On the other hand, it would seem likely that NABS classification would be used at GBARD
level, aimed at indicating the policy objective of the funding streams. However,
attributing NABS classification at this level resulted as unfeasible for most countries
under examination. In fact, any NABS classification observed is typically the result of
estimation made by NSAs, based on their observations of GBARD as whole. The ability to
classify all funding streams is limited to few countries, and a large majority can only
attribute the funding to categories 12 (General advancement of knowledge-R&D financed
by General University Funds GUF) or 13 (General advancement of knowledge- R&D
financed by other sources than GUF).
A final limitation concerns lack of information on the breakdown of funding instruments
by beneficiaries, a problem already familiar from other projects, such as JOREP. For
some countries, it is difficult to disentangle the amounts of funding allocations devoted to
different types of beneficiaries, because it is a matter that can only be analysed at the
execution level. This means that the availability of statistics from different countries is
uneven.
The limitations described above do not affect the comparability of the data, which is very
good, and consequently the robustness of the analysis presented in the report. For more
-
26
information on the characteristics of the data collected in the different countries the
reader can also refer to the Data Quality Report and to Section 2 of the Public Funding
Country Profiles.
3.4 Data on EU structural funding
EU structural funding is an important source for the R&I budgets within the individual
member states. The intention of the PREF project was to obtain data on structural
funding directly from the EU, for a full overview of funding arriving in the different
countries over the years, however after determined attempts, our observation is that it
seems impossible to receive such data. In January 2015, the Commission provided the
consortium a link to a DG REGIO database with information on planned financing under
the different ESI Funds for 2014-2020. The data was provided as a financial table and
broken down by fund, program, priority axis, thematic objective and category of region
(more or less developed, etc.) However, these data refer to planned and future
allocations, which are not directly useful for PREF. Furthermore, the historic and current
allocations of structural funds have also been determined below the EU level, based on
priorities and considerations. Therefore, the PREF project had to proceed from nationally
reported data on past use of SFs, developing these in a manner suited to the study aims
and analyses. The problems in deciphering EU funding descended to the analyses at
national level. For many eastern countries, structural funds constitute a substantial share
of total funding. Since internal ministries and other public funding agencies, often
combined with national funding, allocate SFs, it is difficult to disentangle the shares of
state and structural funding at the stream and instrument levels. In this case, data have
been included in the dataset only when contacts at NSAs were able to assist in solving
the problem.
Figure 2.3 below shows the volume of EU funding for research and innovation in EU
member states (planned amounts for period 2014-2020), and reveals the important role
of structural funds in some countries. However, the data cannot be further commented,
since they include funding for both R&D and innovation and no possibility to separate
them.
Figure 3.3 European Structural and Investment Funds for Research and Innovation projected for EU member states (2014-2020)
Source : EU cohesion data available at : https://cohesiondata.ec.europa.eu. Unit: million Euro.
-
27
4 Intensity of public funding for R&D in European countries
This section diagrams the evolution of intensity of national public R&D funding over the
years considered in the PREF study. The analysis is based on EUROSTAT/OECD data4 and
information from PREF data collection on funding streams.
This section of the report summarises the changing intensity of public funding, including
distribution to public and private sector, and the distribution to fields of higher education
(FORD classification). Section 4 will delve more deeply into issues of institutional versus
project funding, policy objectives and beneficiaries (Indicators 9.1.1 and 9.1.2). The
different analyses also provide evidence of congruence between official statistics and
PREF data.
4.1 Evolution of intensity of public R&D funding, 2000-2014
According to the Frascati Manual (OECD, 2015), “government budget allocations for
R&D” (GBARD) is a measure of the public investment in R&D based on analysis of the
public budget. The indicator relies on identification of the specific budgetary lines
intended for R&D, such as transfers to RFOs or research organisations. The indicator is
considered particularly suited to revealing policy intentions behind the mobilisation of
public funding for research, and the policy objectives pursued by different governments
using R&D allocations.
National funding for R&D has changed through the first years of the 2000s due to general
reductions in the public budgets in times of economic crisis, and through modification in
modes of allocation. Almost all European governments have reduced R&D funding,
except for a few cases showing strong counter-cyclical behaviours, generally aimed at
stimulating economic growth and encouraging private investment in R&D. Figure 3.1
shows GBARD for the years 2000 and 2014 in the 40 countries covered under the PREF
project, as percentage of general government expenditures. This indicator reveals that
funding allocations have grown in different countries, mobilizing a higher share of public
expenditures than at the opening of the period examined.
We observe a group of countries in western Europe with positive evolution of public
investment, where government expenditures became much more important over 2000-
2014. These include Germany, Switzerland, Austria, Denmark, Luxembourg, Portugal and
the nordic countries, except Finland, which shows a strong reduction in this indicator.
Other western European countries instead show reductions in GBARD as percentage of
state expenditures, in some cases quite substantial (e.g. France, Spain, Italy, UK and
Netherlands). Eastern countries generally show positive trends in percentages of public
funding for R&D; for the Czech Republic, Estonia, and Slovakia this is particularly strong.
We also observe that average annual real growth of GBARD during the 2008-2011
economic crisis shows highly disparate trends among European countries, with strong to
very strong negative values in NL (-0.9%), BE (-1.1), FR (-1.3%), UK (-3.7%), IT (-
5.2%) and ES (-5.3%), contrasting with strong to very strong growth in some other
countries (FI +0.6%, SE +2.8%; DK and AT +3.6%, DE +4.1%). Apart from these
specific countries, the remainder of the European nations observed show negative values
(-1.4% on average) (IUC, 2013).
4 EUROSTAT, release of April 2016.
-
28
Figure 4.1 GBARD as percentage of general government expenditures: PREF countries (2000/2014)
Source: EUROSTAT. Note. Start year is 2001 for EU28; 2004 for CY and MT; 2005 for HU; 2008 for HR, TR, JP, US. End year is 2011
for TR; 2013 for JP, US. Data for CN, FYROM, IL, LI, ME, RS are not available.
Fig. 4.2 presents the absolute values of total GBARD in the PREF countries, plus the total
GBARD of EU-28 in 2000 and 2014 (in Euro, current prices).
We can immediately note two meaningful aspects in evolution of public R&D funding.
First, growth in the GBARD of the EU-20 countries from 2000 to 2014 (current prices)
pushed the European area towards an absolute value of public investment close to the
total GBARD for the USA. However, even considering all EU-28 countries, factoring in the
data from Fig. 4.1, the overall investment in GBARD did not fully achieve US levels of
public funding, despite the policies aimed at 3% targets which were to be adopted by
countries as a consequence of the Lisbon strategy. Second, the unique position of a small
-
29
number of countries within the EU becomes very clear: those with very high values of
GBARD in 2000 actually reinforced this position over the interval to 2014, showing the
highest increases in this indicator. The most striking examples are Germany, among
larger European countries, and Switzerland among smaller ones. The counter-cyclical
policy on public R&D investment in these countries, observed in the previous section, is
confirmed. Austria, Denmark, Sweden and Norway also continued strong policies on
public investment, although for these countries the growth was not as great. If we
consider the values of the other countries in combination with the evidence from Figure
4.1 we can conclude that in many cases, growth of GBARD fom 2000 to 2014 has been
weak.
Figure 4.2 Total GBARD: PREF countries and US (2000/2014)
Source : EUROSTAT. Unit : million EUR (current prices). Note. Start year is 2001 for EU28; 2004 for CY and MT; 2005 for HU; 2008 for HR, TR, JP, US. End year is 2011
for TR; 2013 for JP, US. Data for CN, FYROM, LI, IL, ME, RS are not available.
-
30
Fig. 4.3 shows distribution of GBARD by policy objectives, classified according to
Nomenclature for the Analysis and Comparison of Scientific Programmes and Budgets
(NABS), for indication of the policy intentions of R&D funding.
The figure illustrates that the relative importance of different policy objectives has
remained very stable over 2000-2014, with exceptions. In particular, a significant
reduction in funding for Defence R&D (NABS 14) has been matched by an overall
increase in funding for “general advancement of knowledge”, implemented through both
general university funding (NABS 12) and other channels (NABS 13).
Figure 4.3 GBARD classified by NABS: EU-28 countries (2007/2014)
Source : EUROSTAT. Unit : million EUR, current prices. NABS 01: Earth, NABS 02: Environment; NABS 03: Space; NABS 04: Transport, telecommunication and other
infrastructures; NABS 05: Energy; NABS 06: Industrial production and technology; NABS 07: Health; NABS 08: Agriculture; NABS 09: Education; NABS 10: Culture, recreation, religion and mass media; NABS 11: Political and social systems, structures and processes; NABS 12: General advancement of knowledge: R&D financed
from General University Funds; NABS 13: General advancement of knowledge: R&D financed from other sources than GUF